Apidays New York 2024 - The value of a flexible API Management solution for O...
Alternatives to Agile Estimation
1. $ $
$
Alternatives to
Agile Estimation
Neil Killick
@neil_killick #NoEstimates
neilkillick.com
2. #NoEstimates
"As a software practitioner...
I want to shift focus onto delivering value to
stakeholders & customers and away from
estimating...
So that I am empowered to make the right
decisions & build the right thing"
#NoEstimates is not refusing to estimate.
It's about asking if there is a better way.
3. How long will this project take?
How much will it cost?
Should we go ahead?
18. Slice n' Dice
Small = Simple, Unambiguous
1. Are there >1 acceptance criteria?
2. Are there >1 use cases or user scenarios?
3. What is the simplest technical solution (e.g.
"hamburger method")
19. Why not story points?
Release scope
going up?
Risky to say "not small is OK"
Shifts focus from value to estimated cost
Getting better at estimating delivers no value
Velocity metric gets abused
23. Visualise Cycle Time
Backlog
(options) R1 Ready In Progress Done
3 wks
2 wks
3
+
4 5
Cycle time = 7/5 = 1.5 weeks
4 wks Help PO understand the
devastating impact of WIP
24. Constraints force Decisions
Backlog
(options) R1 Ready In Progress Done
3 wks
?
2 wks
?
● Upcoming stories usually smaller
● Work not done is only "options"
4 wks ● Until PO decides to implement, size
doesn't matter
25. We value
Delivering over Estimating
● Tech infrastructure must
support rapid delivery
● Capability crucial, despite
biz reluctance to release
● Continuous delivery
reduces "need" for
DevOps
estimates
26. Towards #NoEstimates
● Small releases;
Minimum features
● Don't accept >1
point; Slice n' Dice
● Shorten delivery
cycles; Decisions
● Don't rely on
estimates; Measure
27. #NoEstimates
Embrace the delicious uncertainty of
software design and development
Make the uncertain... certain
(delivery dates & cost)
Real constraints inspire
creative solutions to problems
Ask yourself why you estimate and
what you use the estimates for
28. Questions?
Hands up!
Who here estimates?
Who chooses to estimate?
Who doesn't estimate?
● Portfolio level estimation
● 5 potentially valuable projects,
only 1 team
● How would you proceed without
estimates?
● Volunteers to start